Overview of the Third Generation Mobile Communications

1
Overview of the Third Generation Mobile
Communications

• Xiaobo Zhou

2
Contents

• Progress of the Mobile Communication
• Standards WCDMA, CDMA-2000 phase II, TD-SCDMA
  and TD-CDMA
• Introduction to the Mobile Communication Basic
  Concepts, Key Technologies
• Wireless Channel Estimation

3
Progress of the Mobile Communication

• The first generation cellular mobile
  communications (1980 )
• Technology FDMA and Analog Technology.
• Systems AMPS(USA), NMT-900(Sweden), HCMTS(Japan)
• Shortages Only lower Frequency available, same
  Frequency band Inference, poor Security.
• Advantagesconvenience for communication

4
Progress of the Mobile Communication

• The second generation cellular mobile
  communications (1992)
• Technology TDMA, TDMA hybrid FDMA
• Systems DAMPS(USA, IS-54), GSM
• Advantages Higher Frequency available, good
  Security, higher Capacity, good speech QoS

5
Progress of the Mobile Communication

• Technology CDMA(Qualcomm)
• Systems CDMA(IS-95)
• Advantages Higher Frequency available, good
  Security, Soft Capacity, Higher Capacity, Speech
  Activity Technology, Diversity Technology.
• ShortagesFocusing on Speech Service, lower rate
  Service, the Capacity does still not satisfy the
  ever-growing demand, no Multimedia Service and no
  higher rate Service

6
Progress of the Mobile Communication

• The third generation 3G (1996-2005)
• High mobile velocity(300-500km/hour) less than
  100km/hour(GSM)
• To carry out the global wandering tour District
  and Country(GSM)
• Support Multimedia Service, especially Internet
  Service, 144kb/s(Outdoor and higher velocity ),
  384kb/s(from Outdoor to indoor, lower velocity),
  2Mb/s(indoor) Speech of QoS and other services
  4-100-200kbs/s(GSM, lower velocity)

7
Progress of the Mobile Communication

• Convenience for transition and evolvement or
  innovation, compatibility with networks
• Highest spectrum availability, higher QoS, Speech
  Recognition Technology, lower Cost, higher
  Security
• Advantage technologies such as Adversity
  transmitting and receiving, Multipath Combining,
  Turbo Code, Channel Estimation, SIR measurement
  and TPC, Space-time technology, Multiuser
  Detection and Interference Cancellation,
  Beamforming and Smart Antennas, Soft handoff
• The 3Gs aim is to implement truly anybody at any
  place to communicate with anyone at any time

8
Standards WCDMA, CDMA-2000, TD-SCDMA,TD-CDMA

• FPLMTS in 1986. IMT-2000 in 1996
• Japan WCDMA to be central technology in 1997.
• EuropeWCDMA (FDD model) and TD-CDMA(TDD).
• America CDMA-2000 Phase II
• China TD-SCDMA(1998)
• The Standards of 3G

9
WCDMA

• Ompany NTT, NEC, Nokia, EricssonSmallest
  bandwidth 5MhzTechnology DS-DCDMAWorking
  type FDD/TDD Chip rate 4.096McpsLength of
  frame 10/20/30msSynchronous Synchronous and
  asynchronousModulation QPSK/BPSKChannel
  structure of the inverse link Pilot/TPC/Dedicated
  Channel code Convolutional codingTurbo
  codingEnhanced technique Multiuser
  detection(O)/Smart Antennas(O)Handoff Soft
  handoff(complicated)TPC Fast TPC(1600)Forward
  and inverse Detection Pilot assistantSpeech
  code Variant rate

10
CDMA-2000 Phase II

• Company Qualcomm, Motorola, Lucent,
  NortelSmallest bandwidth 31.25MhzTechnology D
  S-CDMA and multicarrierWorking type FDD Chip
  rate 1.2288/3.686McpsLength of frame
  10/20/30msSynchronous Synchronous and
  asynchronousModulation QPSK/BPSKChannel
  structure of the inverse link Pilot/ControlChann
  el code Convolutional codingTurbo
  codingEnhanced technique Multiuser
  detection(O)/Smart Antennas(O)Handoff Soft
  handoff(IS-95)TPC TPC(1800)Forward and inverse
  Detection Pilot assistantSpeech code Variant
  rate

11
TD-SCDMA, TD-CDMA

• Company DATANG, HUAWEI, ZHONGXIN,(
  Nortel)Smallest bandwidth 1.25MhzTechnology TD
  MA and multicarrierWorking type TDD Chip
  rate McpsLength of frame 10/20/30msSynchronou
  s Synchronous and synchronousModulation QPSK/B
  PSKChannel structure of the inverse link
  Dedicated/Pilot/Dedicated Channel code
  Convolutional codingTurbo coding(O)Key
  technology Joint detection, Smart
  AntennasHandoff Soft handoff(?)TPC
  Forward and inverse Detection Pilot
  assistantSpeech code Variant rate

12
Key Technologies

• 1, Auto Frequency Control(AFC), Acquisition or
  Cell Searching, Chip Tracking, Auto Gain
  Control(AGC), Multipath Searching, Root Cosine
  Filters Design, Transmit Diversity, Mobile
  Location, Speech coding. Linear Power
  Amplifier(LPA), Adaptive Echo Cancellation.
• 2, Channel Model Simulation
• 3, Channel Estimation
• 4, Channel Code, especially Turbo Code

13
Key Technologies

• 5, Diversity receiver and RAKE combiner6,
  Multiuser Detection and Interference
  Cancellation7, Joint Detection8, Smart
  antennas9, SIR measurement and TPC 10, Soft
  handoffs in CDMA Mobile systems11, Multimedia
  Communications in 3G12, Speech recognition
  technology in 3G13, Security algorithms14,
  Route technology and Packet access

14
Basic Concepts

• Gaussian channel and multiuser receiver
• A baseband digital direct sequence(DS)CDMA
  network of K users. The received signal can be
  modeled as
• where 2M1 is the number of data symbols per user
  per frame, T is the symbol interval, is a
  collection of independent equiprobable random
  variables,and the user signaling waveforms are
  of the form
  
  
• is a signature sequence of s
  assigned the kth user and is a normorlized
  chip waveform of duration , where .

15

• Rayleigh channel and a single user receiverIn
  the case of the time-multiplexed pilot channel,
  the QPSK symbol sequence is mapped over a
  sequence of slots, each containing data
  symbols preceded by pilot symbols placed at
  the beginning of each slot. The resultant symbol
  sequence is spread over much wider bandwidth by a
  spreading sequence. The slot length
  , where T is the QPSK symbol duration. In the
  case of the parallel pilot channel, on the other
  hand, data and pilot channel are spread by
  orthogonal spreading sequence. Assuming that the
  multiple channel has resolvable,
  frequency-nonselective path, the spread signal
  received over a multipath channel can be
  represented as
• and are the complex-valued channel
  gain and time delay of the l-th
  path(l0,1,,L-1), respectively, and s(t) is the
  transmitted spread signal waveform. We assume

16
(No Transcript)
17
(No Transcript)
18

• Diversity receiver and RAKE combiner
• A diversity scheme is a method that is used to
  develop information from several signals
  transmitted over independently fading paths. The
  objective is to combine the multiple signals and
  reduce the effect of excessively deep fades.
  Diversity schemes can minimize the effects of
  fading, since deep fades seldom occur
  simultaneosly during the same time intervals on
  two or more paths.
•  Since the chance of having two deep fades from
  two uncorrelated signals at any instant is rare,
  the effect of the fades can be reduced by
  combining them. There are two general types of
  diversity schemes. One is called the macroscopic
  diversity scheme. The macroscopic diversity
  scheme is used for combining two or more
  long-term lognormal signals, which are obtained
  via independently fading paths received from two
  or more different antennas at different
  base-station sites. The microscopic diversity
  schemes is used for combining two or more
  short-term Rayleigh signals, which are obtained
  via independently fading paths received from two
  or more different antennas but only one receiving
  cosite.
• The macroscopic diversity schemeapplied on
  different-Sited Antennas.
•  
• The microscopic diversity schemes. At the base
  station and at the mobile unit, there are space
  diversity, Polarization diversity, Angle
  diversity, frequency diversity, time diversity.

19
(No Transcript)
20
(No Transcript)
21

• Rayleigh channel model(ITU-M R.1225)A central
  factor of mobile radio propagation environments
  is multipath propagation causing fading and
  channel time dispersion. The fading
  characteristics vary with propagation environment
  and its impact on the communication quality is
  highly dependent on the speed of the mobile
  relative to the serving base station. The purpose
  of the test environment is to challenge the RTTs.
  
•  
• The key parameters to describe each
  propagation model would include
• --time delay-spread, its structure, and
  probability distribution of time delay spread.
• --geometrical path loss rule(e.g. and excess
  path loss
• --shadow fading
• --multipath fading characteristics(Doppler
  spectrum) for the envelop of channels
• --operating radio frequency

22

• 1, Long-term Fading
• Long-term fading components, which
  contribute only on to propagation-path loss must
  be removed. We must estimate the local mean. in
  general, the local mean can be obtained by
  measurement.
•  
• 2, Short-term Fading
• 3, Frequency-flat fading and Frequency-selective
  fading
• Frequency-flat fading channel is composed
  of long-term fading and short-term fading.
  Frequency-selective fading channel is composed
  multipath channels with different time delay
  spread, which each channel is Frequency-flat
  fading channel.

23
(No Transcript)
24

• Channel Model Simulation
• There are mainly two methods of channel model
  simulation. The first method is filters design.
  The Classical Doppler spectrum is
• The second method is harmonic decomposition
  technique. The simplest nondegenerate class of
  process which exhibits uncorrelated
  dispersiveness in time delay and Doppler shifts
  is known as the Wide Sense Stationary
  Uncorrelated Scattering(WSSUS)
• Now, we use a series of the form
  
• can be find, approximating
  arbitrarily close, in the mean square sense,
  provided we take small enough. The
  are uncorrelated complex random variables with
  zero mean and variance

25
Wireless Channel Estimation

• Wienner filter
• Gaussian Interpolation
• WMSA(Weighted multi-slot averaging)
• Polynomial Interpolation
• Adaptive Lattice Weighted Algorithm
• Kalmann Filter based on AR model
• Joint Data Message to Estimate Channel
• Channel Estimation based on Fuzzy systems
• Blind Channel Estimation

26
Wienner filter

• Fading Channel output
  , Complex channel gain c(t) is
  time-variable, which auto correlation is
  . In the
  receiver, the received signal pass through a bank
  of matched filters. The matched filters output is
• In generality, we assume b(0) to be a pilot
  symbol. Now, we want to detect symbols
  , Then, the linear estimator
  is
• According to Wienner filter, we can obtain
  the optimal h(k).

27
WMSA

• simple Coherent RAKE receiver
• The channel estimation filter is to estimate the
  value of using the pilot channel and
  its estimate is denoted by L
  despread and resolved signal components are
  multiplied by the complex conjugates of
  s before combined(maximal-ratio
  combing(MRC)). The RAKE combiner output at the
  m-th symbol position of n-th slot is therefore,
  represented as

,
28

• Finally, the RAKE combiner output is
  de-interleaved and soft-decision Viterbi decoded
  to recover the transmitted data.
• We assume the channel estimator to be which
  denotes the channel estimator at the m-th symbol
  position of n-th slot associated with the l-th
  path . In the case of fading, we can extend the
  observation interval to several slots and
  coherently add several consecutive instantaneous
  channel estimates to further increasing
  SNR Therefore, the instantaneous channel
  estimates need to be smoothed by a smoothing
  filter. The smoothing filter is expressed as
• where is the pilot channel estimator at the
  n-th slot. is the coefficients of the
  filter

29

• As
• AS
• The RAKE combiner (Max-ratio Combing(MRC) output
  at the m-th symbol position of n-th slot is
  therefore, represented as

30
(No Transcript)
31
Kalmann Filter based on AR model

• the state process and observation process of
  Kalman filter.
• The traditional transition matrix
• The traditional Kalman filter can only be used
  for the estimate of slow fading channel, which
  channel gain remains almost same in one slot. The
  kernel of this paper demonstrates how to get this
  transfer probability matrix. We get this matrix
  through two methods. AR model is used in the
  state process of Kalman question. To get AR
  model's coefficients, we apply two methods, one
  of which is adaptive LMS algorithm, and the other
  of which is Durbin retrieve method.

32
(No Transcript)
33
(No Transcript)
34
(No Transcript)